Microsoft Machine Learning Server in Agriculture: How the Fourth Industrial Revolution is Driving the Second Green Revolution

Microsoft Machine Learning Server in Agriculture

Back in the 1940s, American scientist Norman Borlaug triggered the start of the Green Revolution. He first started his research initiatives in Mexico, developing new varieties of wheat.

These were high-yield, disease-resistant varieties, which together with new agricultural technology led to a massive rise in production. Mexico soon had more wheat than its citizens needed and became a top global exporter.

The success of these initiatives spread globally in the 1950’s and 1960’s. Unfortunately, it was short-lived. In spite of their benefits, the practices were unsustainable and led to the rise of new issues.

Today, mankind is in the throes of the second green revolution. More and more farmers are embracing technology in an attempt to use less to feed more. Many are also keen on limiting the negative impacts of agricultural practices on the environment and human health.

Microsoft’s Machine Learning Server is playing an active role in making the fourth industrial revolution a major driving force of the current green revolution. Here are some of the ways in which this is happening:

Machine Learning-Based Precision Farming Technology

Farmers are making extensive use of a combination of artificial intelligence and machine learning to enhance agricultural practices. For instance, they are using wireless sensors, robots and drones to collect data on plants’ growing conditions.

They then use machine learning cloud services to process data and interpret findings. Based on this, they can adjust crop input so as to boost yields.

Thanks to the use of this technology, they can identify the optimal sowing dates, precise allocation of water and fertilizer as well as identify crop diseases in time. They can also make production and weather forecasts, pricing predictions and other input to determine how much to sow.

As is the case with all machine learning-based models, the systems in use become smarter with time, augmenting farmers’ abilities and fostering sustainable production.

But in most cases, the areas where farmers need technology the most lack the infrastructure to support it. Thus, Microsoft launched a project known as the Airband initiative to address this need.

The initiative provides rural communities in Kenya, South Africa, Colombia, India and the US with affordable internet. Through this communal connection, farmers can access machine learning-related cloud services and optimize performance.

Another initiative, the Microsoft FarmBeats program, offers an end-to-end platform which aims to increase farming productivity and profitability. It features a combination of machine learning algorithms, low-cost sensors and drones to operate.

With time, the use of these technological tools should help farmers increase yields to meet global needs for food sustainably.

Safeguarding the Food Chain

Besides food production, another major concern for farmers and others in the agricultural sector has to do with contamination. There have been serious challenges in keeping toxins out of the food chain sustainably.

Among the top culprits in this regard are mycotoxins such as aflatoxin, which is highly carcinogenic. Detecting these toxins is extremely difficult as they are not visible to the human eye and neither can you taste or smell them.

To make matters worse, cooking food which has the contaminant will not make it safe and it is heat-resistant. And unfortunately, the identification of just two grains in 10,000 is all it takes to render an entire lot unfit for human consumption.

In developed countries, the concern is mostly economic as a producer cannot sell contaminated food. But in developing countries, it is a major health problem as it may mean eating contaminated food or going hungry.

Consumers have no way of telling whether or not their food has the contaminant. So it is the harvesters’ and processors’ responsibility to do so.

A processing company known as Buhler is using a combination of machine learning, camera and UV lighting to address this challenge. They have created an optical sorting machine, LumoVision, which photographs individual kernels and separates contaminated maize grains.

At the onset, the project was not financially viable. But with the use of Microsoft’s machine learning server, the team was able to design a cost-effective model. It identifies maize grains with mold contamination and removes them using high precision air ejection.

Though the technology only works with maize at present, it underscores the possibilities that Microsoft Machine Learning Server users in agriculture can access.

Considering the amount of food that goes to waste due to substandard processing in underdeveloped countries, the economic benefits would be massive. At the same time, such an initiative can save lives, eliminating the fatal effects of chronic exposure to such toxins.

Streamlining the Online Food Shopping Experience

Another innovative use case of Microsoft Machine Learning Server within the food industry has to do with predictive customer service. The food industry is highly competitive and one of the ways to differentiate a business is by connecting better with the customer. How can technology make this possible?

Today, there are a significant number of food delivery services operating online. Apart from offering convenience to the customer, they also have another key benefit. They get access to tons of data courtesy of customer orders and activities on their web pages.

Customers do not want to spend inordinate amounts of time online placing orders. With this in mind, the use of a machine learning-based system can make it possible to anticipate their needs.

A major challenge in putting this theory into practice is that customers vary remarkably and place different orders on each purchase. Therefore, for such a system to succeed, it has to be relevant to the customer during every order.

With a machine learning system, this is possible as it automatically analyzes data and identifies patterns with ease. To illustrate, it would easily recognize that a certain customer buys grains once a month, vegetables weekly and cooking oil every other month. This would make it possible to predict the timing for the next vegetable purchase and so on.

Based on customer behavior, it would be possible to train a machine learning model and offer predictions to enhance customer experience. A UK company known as JJ Services uses such a model on its online food delivery service.

When a repeat customer logs in or places a call, the system automatically fills the order pad. As a result, customers spend less time shopping. Additionally, the system creates opportunities for cross-selling. To do this, it reviews orders to determine whether the items on a shopping list indicate a likely need for other products.

With this combination of predictive analysis and recommendations, such a machine learning model is a win-win for both sides. The customers get what they need in one order and enjoy the personal touch while the seller gets more sales through recommendations and reaps the benefits of high customer satisfaction.

Microsoft Machine Learning Server – From Farm to Fork

The agricultural sector is among the highest impact industries worldwide. In addition to employing 30% of the global population, it contributes 10 to 30% of the world’s GDP, with $4.8 trillion global revenue.

However, it also consumes 70% of the world’s water resources and is responsible for 24% of global greenhouse emissions. Any technology that can enhance the positives and minimize the ills of this sector is crucial to mankind’s survival in more ways than one.

The use of Microsoft’s Machine Learning Server in the agriculture sector offers clear benefits on farming, food processing and even delivery as the above examples illustrate.

Do you use SharePoint? Try our toolkit
Download SharePoint Essentials Toolkit Now
Download the SharePoint Essentials Toolkit
Janica San Juan

Leave a Reply

Your email address will not be published. Required fields are marked *

This site uses Akismet to reduce spam. Learn how your comment data is processed.